Thrombospondin-1 and CD47 signaling regulate healing of thermal injury in mice

Abstract

More than 2.5 million Americans suffer from burn injuries annually, and burn management is a major public health problem. Treatments have been developed to manage wound injuries employing skin grafts, various dressings and topical and systemic agents. However, these often achieve limited degrees of success. We previously reported that targeting the interaction of thrombospondin-1 with its signaling receptor CD47 or deletion of the genes encoding either of these proteins in mice improves recovery from soft tissue ischemic injuries as well as tissue injuries caused by ionizing radiation. We now demonstrate that the absence of CD47 improves the rate of wound closure for a focal dermal second-degree thermal injury, whereas lack of thrombospondin-1 initially delays wound closure compared to healing in wild type mice. Doppler analysis of the wounded area showed increased blood flow in both CD47 and thrombospondin-1 null mice. Accelerated wound closure in the CD47 null mice was associated with increased fibrosis as demonstrated by a 4-fold increase in collagen fraction. Wound tissue of CD47 null mice showed increased thrombospondin-1 mRNA and protein expression and TGF-β1 mRNA levels. Activation of latent TGF-β1 was increased in thermally injured CD47-null tissue as assessed by phosphorylation of the TGF-β1 receptor-regulated transcription factors SMAD-2 and -3. Overall these results indicate that targeting CD47 may improve the speed of healing thermal injuries, but some level of CD47 expression may be required to limit the long term TGF-β1-dependent fibrosis of these wounds.

Keywords: CD47; TGF-β1; Thrombospondin-1; Wound healing.

Fig. 1. Role of the TSP1/CD47 axis in thermal wound healing

Thermal wounds were inflicted in the dorsal area of WT, CD47−/− and TSP1−/− mice, left panel (A) shows representative pictures of wound healing. The size of each wound was measured using a caliper. Area was calculated using the formula A = πa (B). Area and percent wound closure (C) were normalized from WT wounded measurements. N = 6 *p < 0.05.

Fig. 2. CD47 deficiency regulates collagen deposition to improve wound healing

Skin sections of tissue harvested at day 7 and 17 were stained with picrosirius red to detect collagen fibers denoted by arrows (A). Collagen deposition was quantified by a computer-assisted counting technique with a pixel counter at day 7 (B) and day 17 (C) N = 3 *p < 0.05.

Fig. 3. CD47 and TSP1 differentially regulate TGF-β1 expression and processing in thermal wounds

Paraffin embedded wounded skin tissue sections were harvested at day 7 and stained using an antibody to TGF-β1 (A). Expression of the propeptide and mature form of TGF-β1 protein was measured in tissue extracts from wounded skin by western blotting (B). RNA was isolated from wounded skin from all three genotypes, TSP1−/− mice were injected with antisense morpholino to CD47 (CD47M) and TGF-β1 expression was determined by RT-PCR (C). Expression was normalized to that in injured WT tissue (mean ± SD, N = 3 *p < 0.05).

Fig. 4. CD47 deficiency increases functional TGF-β activity as assessed by SMAD signaling

Wounded tissue sections were paraffin embedded and stained using specific antibodies to phosphorylated SMAD2, phosphorylated SMAD3, and SMAD4 (A) brown stain denotes immunoreactivity (arrows). Tissue sections were scored under light microscope using cell counter mode with ImageJ software, and the percentages of cells positive for p-SMAD2 (B), p-SMAD3 (C), and SMAD4 (D) are presented as mean ± SD, N = 4.

Fig. 5. TSP1 expression is up-regulated in injured CD47 deficient skin tissue in a TGF-β1-dependent manner

Wounded tissue sections were paraffin embedded and stained with TSP1 antibody A6.1 (A). Arrows denote positive staining in dermal epithelium. mRNA was isolated from wounded tissue extracts and subjected to quantitative RT-PCR using specific primers for TSP1. Expression is normalized to that in injured WT tissue (mean ± SD, N = 3 *p < 0.05) (B). Wounded CD47−/− mice were injected with antisense TGF-β1 morpholino (TGF-β1M) or mismatched control morpholino (TGF-β1MM), and tissue was harvested 24 h later for western analysis of TGF-β1 expression (C, upper panel) and immunohistochemical analysis of TSP1 expression (C, lower panel).

Fig. 6. Activation of TGF-β1 signaling in CD47 deficient conditions is dependent on TSP1 expression

Wounded TSP1 null mice were injected I.P. with CD47 morpholino (CD47 M) at day 5, and tissue was harvested at day 7. Efficiency of the CD47 morpholino was confirmed by western blotting using murine CD47 antibody clone map 301 (A). Tissue sections were par-affin embedded and stained with specific antibodies to TGF-β1, p-SMAD2, p-SMAD3 and SMAD4 (B). Arrows denote immunoreactivity by brown stain.

Fig. 7. CD47 deficiency increases blood flow and VEGF/VEGFR2 signaling to improve wound healing

Seven days after inflicting thermal wounds, skin was excised, and paraffin embedded sections were stained using an antibody to CD31 (A). Endothelial vessel density was determined by counting vessel “hot spots” delineated by the CD31 stained cells observed at 20× magnification under light microscopy (B). One week after injury, mice were subjected to laser Doppler imaging to measure blood flow, and representative images are presented (C). mRNA was isolated from wounded tissue extracts and subjected to quantitative RT-PCR using specific primers to VEGF and VEGFR2, using mouse actin as a gene control (D and F).

Fig. 8. Deficiency of CD47 regulates inflammatory response in healing wounds

Wounded skin sections at 48 h and at one week post wound infliction were stained using hematoxylin and eosin to determine tissue structure (A). Polymorphonuclear cells present in epithelium are denoted by arrows and were quantified under light microscopy(B). N = 3 *p < 0.05.